|Publication number||US4298954 A|
|Application number||US 06/034,234|
|Publication date||Nov 3, 1981|
|Filing date||Apr 30, 1979|
|Priority date||Apr 30, 1979|
|Also published as||CA1141476A1, DE3064204D1, EP0018518A1, EP0018518B1|
|Publication number||034234, 06034234, US 4298954 A, US 4298954A, US-A-4298954, US4298954 A, US4298954A|
|Inventors||George A. Bigelow, Ted A. Rehage, Frankie S. Shook|
|Original Assignee||International Business Machines Corporation|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (17), Non-Patent Citations (1), Referenced by (71), Classifications (8)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The present invention relates to buffering data signals and more particularly to providing a wideband buffer system.
Buffering data signals between units or systems having different data rates for rate changing is well known. A particular version of such buffering for enhanced buffering functions includes alternating between two buffer units. For example, while a first buffer unit is receiving signals to be buffered, a second buffer unit is supplying signals that were previously received. When the one buffer is empty and the other buffer is full, then the functions are alternated between the buffer units. Such a system is shown in Garrison et al, U.S. Pat. No. 3,020,525.
In any buffering application, it is desired to have as wide a bandwidth of data transfer as possible at lowest cost. A wide data bandwidth can be achieved by making the buffer larger. However, when a buffer is larger, the cost is increased. Therefore, it is desired to have a relatively wideband buffer for a minimal size buffer unit such that maximal performance can be achieved with minimal cost.
Alternating buffer apparatus constructed in accordance with the invention includes a plurality of register sets with switching means for alternatively accessing the register sets for receiving and delivering data signals. Addressing means indicate when the various register sets have data signals in predetermined ones of the registers. Each of the register sets also include ready means indicating that a predetermined number of registers have data signals for indicating their readiness to switch whenever another register set is empty; such that the empty register set can receive signals while the switching register set can deliver signals before it is full, thereby increasing the data bandwidth.
A plural ported apparatus incorporates a plurality of such alternating data buffer sets such that various inputs to a high speed unit via the wideband buffer sets to enhance data transfer through the plurality of ports for making the high speed unit more accessible to each and every one of the ports. Such apparatus constitutes an improved signal burst accommodating means.
The foregoing and other objects, features, and advantages of the invention will be apparent from the following more particular description of the preferred embodiments of the invention, as illustrated in the accompanying drawings.
FIG. 1 is a diagrammatic showing of a plural ported system incorporating the teachings of the present invention;
FIG. 2 is a diagram of a high speed common unit of FIG. 1 configured as a storage subsystem;
FIG. 3 is a simplified diagrammatic showing of a burst driven port in the FIG. 1 illustrated embodiment and adapted to show principles of the present invention;
FIG. 4 is a logic diagram illustrating operation of a burst driven port of FIG. 1;
FIG. 5 is a flow chart of a burst driven port of FIG. 1; and
FIG. 6 illustrates a connection between the priority circuits and the store of the FIG. 2 illustrated version of a high speed communication unit whereby ports are selected in accordance with a predetermined priority.
Referring now more particularly to the drawing, like numerals indicate like parts and structural features in the various diagrams. In FIG. 1, a high speed common unit 10, which may be a random access memory unit, is connected to a plurality of data burst units, DBU-1, etc., via a like number of burst driven ports, BDP-1 et al. Such data burst units can include magnetic tape units, digital computers, slower speed random access memories, communication lines and the like. It is desired that the operation of the high speed common unit (HSCU) 10 be such that each of the DBU's have an apparent exclusive access to the HSCU 10 even though other DBU's are using it. This apparent exclusivity is achieved by providing a burst driven port BDP between each of the data burst units DBU and HSCU 10.
As an example, FIG. 2 a data burst unit DBU includes a storage control unit (SCU) attached to a storage unit (SU). A storage unit SU can be a magnetic tape or disk unit, for example, which will include a storage media assembly (SMA). In general, such devices when transferring signals require a long burst of signals for maximal efficiency. Intermittent operations require interleaved time consuming mechanical motions which reduce unit efficiency.
FIG. 3 illustrates a data burst accommodating port BDP for bilateral communication; i.e., data flow is in either direction as selected by means (not shown) external to the port. In the illustrated embodiment, the burst driven port consists of two separate buffers or register sets 12, 13 which are controlled to appear as one logical buffer. These separate buffers 12, 13 accumulate data in preparation for storing it in HSCU 10. This action allows for any number of data burst units to be attached to HSCU 10 as determined by the balance of the data burst unit data rate and the HSCU 10 data rate. In one embodiment, each of the buffers 12, 13 has 16 registers. Each of the registers has a predetermined number of bits for storing data, a smaller number of bits for storing error detection correction bits and an additional bit position for storing a mark bit. The mark bit points to the logical end of a data transfer, i.e., the end of a data block. In any practical embodiment, the number of registers, i.e., the buffer size, is determined by the access time of HSCU 10 as measured by each of the data burst units and the rate at which data signals actually arrive at the burst driven port.
Each of the registers in buffers 12, 13 is accessed via addressing circuits 14 and 15, respectively. Decodes 16 and 17 are respectively attached to the addressing circuits for decoding which registers in buffers 12, 13 have data stored therein for transmission to a receiving unit, such as HSCU 10. The decodes 16,17 in turn supply decoded signals to a control 18 which in turn operates the two switches 20 and 21 for alternating data transfer between buffers 12 and 13 and external units. When buffer 12 is supplying signals to HSCU 10, buffer 13 is receiving signals from a DBU; and vice versa. Alternation is switching the above described functions. In transferring signals from HSCU 10 to a DBU, a similar alternation occurs, as is well known. Each of the switches 20 and 21 is respectively connected to data busses 23 and 24 for transferring signals as shown in one of the port connections of FIG. 1. It is preferred that the number of lines in busses 23, 24 be equal to the number of data bits in each of the registers of register sets 12, 13. The usual error correction circuitry and other control circuitry are not shown for purposes of brevity.
In accordance with the invention, decode circuits 16, 17 detect when receiving register sets 12, 13 have received sufficient data signals without being full to enable switching buffers for increasing the bandwidth of the alternating buffer system. In this regard, each of the buffers 12, 13 is arbitrarily divided into four-register sections; each four-register section constitutes an alternation point for alternating the data transfer directions between buffers 12, 13. That is, at each four register boundary or alternation point decodes 16, 17 can actuate control 18 to determine whether or not the other one of register sets 12, 13 is empty. The buffer empty signal from either decode 16 or 17 is combined with a ready signal from the other decode to actuate control 18. For example, if buffer 13 is empty when buffer 12 is ready to switch, then the buffers 12, 13 are alternated by the actuation of switches 20, 21 by control 18. At this point, buffer 13 then begins to receive data signals while buffer 12 supplies data signals. In the reverse sense, when buffer 13 has received data signals up to a transfer or alternation point which yields a ready signal from decode 17, control 18 then responds to the decode 16 empty signal to actuate switches 20, 21. In this manner, it is not necessary for the receiving buffer, set either 12 or 13, to be completely full before alternation. By alternating before a receiving register set is full, the just received data signals can be supplied sooner, thereby widening the bandwidth of the buffer system. It is to be understood that the invention is not limited to pure alternation between two buffers, such as register sets 12 and 13 but that 3 and 4 or more sets of buffers can be sequentially scanned in accordance with the present invention.
Since the buffers 12 and 13, are independent; simultaneous write and read operations occur. This arrangement enables the buffers 12 and 13 to operate in a longer cycle time than would be required if a time sliced single physical buffer were employed. Therefore, it is seen that reduced access time to data supplied to the buffer system is accomplished by the receiving buffer switching to the supply state before becoming completely filled. This arrangement also allows and compensates for variability in incoming data rates.
While the illustrated embodiment shows four alternation points, no limitation thereto is intended; for example, for maximizing the bandwidth switching, alternation can occur at any register.
FIG. 4 illustrates decodes 16, 17 and control circuit 18 for enabling the dynamically variable alternation of the present invention. In particular, decode 16 is shown in detail, it being understood that decode 17 is similarly constructed but which is shown in abbreviated form. Addressing circuits 14 and 15 are shown with their detail connections to the decode circuits 16 and 17.
Each of the addressing circuits has a write counter 30 which indicates the register for receiving data signals and a read counter 31 which indicates the register from which signals are obtained. In any data transfer operation, either the write counter 30 or the read counter 31 is used in each of the addressing circuits 14 and 15. Control 18, which includes a buffer selection latch 32 selects which counter is to be used. In control 18, register set 12 is designated as buffer A, and buffer 13 is designated as buffer B; latch 32 being in the WA, RB state indicates that data signals will be written into buffers 12 (buffer A) while data signals will be read from buffer B, register set 13.
An active signal on line 33 (WA, RB) enables AND circuit 34 in address circuit 14 to actuate write counter 30 to be stepped each time a write signal is received over line 35. That is, a data burst unit DBU-1 or HSCU 10 supplying signals indicates when a signal is available by actuating write line 35. In this situation, data signals are written into buffers 12 (buffer A) while data signals will be read from buffer B, register set 13. In a bilateral transfer, the signal on line 35 comes from either HSCU 10 or data burst unit DBU. The selection of data transfer direction is beyond the scope of the present description and is well known in the data processing arts. A timed signal on line 35 passes through AND gate 34 to increment write counter 30 to select the next register of register set 12 for receiving the next data signals. Since there is also a read function for buffer B, the line 33 is also connected to address circuits 15 wherein the activating signal activates an AND circuit (not shown) for actuating the read counter (not shown) in address circuits 15. Address circuits 15 are identical to address circuits 14. In address circuit 14, AND circuit 37 receives a timing pulse from the read line 38 which receives a read signal from either HSCU 10 or the data burst unit DBU to actuate read counter 31 to indicate the next register to be accessed. Both counters 30 and 31 and address circuits 14, 15 are resettable by the reset signals on line 39 for setting the counters 30, 31 to a reference state at a start-up of operations.
To write in buffer B (13) and read from buffer A (12) latch 32 is triggered to the opposite state such that an active signal is supplied over line 42 to AND circuit 37 of address circuit 14 and the AND circuit (not shown) of address circuits 15 corresponding to AND circuit 34 of address circuits 14. A disabling signal now appears on line 33 for degating AND circuit 34 and address circuits 15 in the read mode.
The decodes 16 and 17 are identical and each consists of three circuits; a ready circuit 44, an empty circuit 45 and a full circuit 46. The circuits are detailed and explained for decode 16, operation of the decode 17 is identical to decode 16. Each ready circuit 44 and the full circuits 46 operate when the corresponding buffer 12, 13 is receiving signals (being written into) while empty circuits 45 operate when the corresponding buffer is supplying signals (being read).
When buffer 12 is receiving signals, ready circuit 44 receives the signal contents of write counter 30 for determining when buffer 12 has registers filled to an alternation point. As previously mentioned, the selection of the alternation point determines the bandwidth of the buffer. Accordingly, ready circuit 44 is programmable by a single-pole double-throw switch 48 for selecting alternation points every fourth register or every Mth number of registers. M is selected by signals received over cable 49 from a control console (not shown) for example. Assume that programming switch 48 is set to the four state, then each alternation point is at any fourth register. In a sixteen register buffer 12, there are four alternation points. Such alternation points are decoded by AND circuit 50 receiving the bit 0 and bit 1 signals over lines 51 and 52 from write counter 30. The signal on line 53 from switch 48 enables AND circuit 50 to pass the alternation point indication on lines 51 and 52 through OR circuit 54, thence to ready line 55 as a ready signal. Ready line 55 goes to control 18 and is used as will be described. Decode 17, ready circuit 44 supplies a ready signal over the second ready line 55'.
When the number of alternation points is to be changed from four, programming switch 48 is set to the M terminal for selecting compare equal circuit 60 to generate a ready signal on line 61. The line 61 ready signal goes through OR circuit 54 to ready line 55. In this instance all of the bit positions of write counter 30 are supplied to compare equal circuit 60 which also receives the signals supplied over cable 49. When the bit pattern on cable 49 equals the bit pattern of write counter 30, the ready signal on line 61 is generated as is known for compare equal circuits.
The second condition for alternation is when the other buffer 13 supplying signals is empty. It will be noted that the write counter 30 is actuated only when data signals are being written into its corresponding buffer 12. When buffer 12 is supplying the signals, the numerical content of write counter 30 remains the same. Therefore, write counter 30 then indicates the last register position which contains data signals. This indication is supplied over cable 65 to empty circuit 45 and compared with the numerical contents of read counter 31. As signals are transferred from buffer 12, read counter 31 is incremented toward the numerical count in write counter 30. Because of the speed of switching, it is desired to have a one register look-ahead for alternating between the buffers 12 and 13. The +1 circuit 66 receives the numerical contents of read counter 31 over cable 67, adds 1 to such numerical contents and then supplies such contents +1 over cable 68 to compare equal circuit 69. Compare equal circuit 69 also receives write counter 30 signals on line 65. Each time AND circuit 37 supplies an incrementing pulse to read counter 31, that pulse is also supplied over line 70 to actuate compare equal circuit 69. When the signal contents on cable 68 equal the signal contents of write counter 30, compare equal circuit 69 supplies an actuating signal over line 71 to set flip flop 72 to the active condition to supply an empty signal over line 73 to control 18. Similarly, empty circuit 45 in decode 17 supplies a similar empty signal over line 73' to control circuit 18. In another embodiment of the invention, flip flop 72 can be a D-type latch. In such a situation line 70 is then connected to the C input of the D-type latch with the line 71 connected to the D input. Also, compare equal circuit 69 can then be a DC translator for making a comparison between the cable 68 signals and the signal contents of write counter 30.
Control circuit 18 responds to the decodes 16 and 17 to alternate the buffers 12, 13. AND-OR (AO) circuit 80 decodes the ready and empty signals from decodes 16 and 17 for triggering control latch 32 to effect such alternation. The A1 AND input portion of AO 80 supplies an activating signal to trigger latch 32 whenever the line 55 ready signal and the line 73' empty signal are activated. In this situation, the line 33 signal was active and latch 32 should be switched to activate line 42. Returning to activating line 33, the A2 AND input portion of A0 80 responds to the line 73 empty signal and the line 55' ready signal to toggle latch 32. An alternate connection is to connect the A1 AND input portion to the reset input R and to connect the A2 AND input portion directly to set input S of latch 32.
The lines 33 and 42 signals also go to switches 20, 21 for actuating same as is well known in the data processing art. Further, the lines 55 and 55' signals are combined in OR circuit 82 to supply a single ready signal to the switches 20, 21 for priming same to switch whenever an empty signal is received.
The empty latch 72 is reset to remove the ready signal on line 73 when the AND circuit 34 supplies its first pulse to write counter 30; that is, the first data is being written into a register of buffer 12 at the address indicated by write counter 30. Then buffer 12 is no longer empty; the empty signal on line 73 should be erased.
It is possible in any buffering situation that overwrite can occur. This error is prevented by the respective full circuit 46 detecting via address circuit 14 that all registers in a buffer 12, 13 that are receiving data signals are full. Circuit 46 then supplies a stop write signal over line 86 through OR circuit 87 to both the connected units HSCU-10 and DBU. Similarly, full circuit 46 of decode 17 supplies a stop signal over line 86'. Each full circuit 46 includes a compare equal circuit 90. This circuit in decode 16 detects equality between the signal contents of write counter 30, plus one as supplied from write plus 1 circuit 91, and the numercial contents of read counter 31, as received over cable 67, to set stop latch 92 to the active condition. Latch 92 supplies a stop signal over line 86 to stop the writing into buffer 12. Latch 92 is reset whenever AND circuit 37 supplies an activating pulse to increment read counter 31 which then removes the data from the register which would be written into next. Immediately the stop write signal is removed enabling the next data signals to be written into the selected buffer. In this manner, the write counter 30 can track the read counter 31 very closely without the danger of overwriting. The numerical contents of the write and read counters repesented by the signals on cables 65, 65' and 67, 67' are respectively sent to the buffers 12 and 13 for addressing same as is well known. The full circuit 46 of decode 17 operates identically with address circuit 15.
In FIG. 3, lines 35 and 38 are separated into two parts, it being understood that an AND/OR circuit (not shown) combines the signals respectively on the single lines 35 and 38 as shown in FIG. 4.
FIG. 5 shows the logical operations described with respect to the FIG. 4 circuit. In this regard, the control shown in FIG. 4 may be implemented in a programmable logic array, as a sequence of programmed operations in a microcomputer or other programmed logical devices. All such implementations are within the scope of the present invention. The FIG. 5 flowchart, as shown, assumes buffer B has been first loaded with data signals. Start 100 is a sequence entry before each byte transfer. In the flow chart of FIG. 5, after start at 100, the first operation 101 writes into buffer A and reads from buffer B, designated as WARB in block 101. Of course, when first starting buffer B is empty, therefore it is not read out, making step 101 WA. Then the programmed unit (not shown) at 102 determines whether or not buffer A is full. If it is full, it must wait for a buffer B empty indication such as is determined at 103 and sensed at 104. If A is not full, then the programmed unit at 105 determines whether or not data is to be supplied. If not, a wait loop including steps 102 and 105 is entered. However, if data is available, as indicated by the write signal on line 35 at 106, data signals are stored into registers 12 (buffer A). Then the programmed unit (not shown) determines at 107 whether or not buffer B (13) is ready. This action corresponds to the detection of the ready signal on line 55. If the alternate buffer is not ready, then another data signal is entered into buffer A. However, if the ready signal is active at 107, then the programmed unit (not shown) determines at 108 whether or not A is full; if it is full, the A full (AF) flag is set at 109 and step 104 executed. If buffer A is not full, then step 104 is directly executed. The alternation of the buffers precedes from branch step 104 which occurs only after B is ready or B is empty (step 103). If B is empty at 104, then the buffers are alternated by setting write B (WB) and read A (RA) at 110. The steps collectively designated by numerical 120 are identical to steps 101-109 except the operations are complementary as to buffer A, B as can be seen from the inspection of the drawings.
When, as shown in FIG. 6, a high speed common unit 10 is operating with several ports, some priority has to be achieved. A simplified priority results when HSCU 10 is four times as fast as any one of the burst units. The speed difference would be the average speed, not peak speed. Accordingly, each port would have one out of 4 time slices of HSCU 10, as is known and as determined by a priority circuit 130. Circuit 130 can be a time slicer which activates the HSCU interface circuits 131 to control the storage portion 132. The priority circuit 130 receives requests for access to store 132 over access request lines collectively enumerated by number 133, such as from OR circuit 82 of FIG. 4, which are understood to be received from burst driven ports BDP-1 through BDP-4. Other forms of priority circuits can be used with equal efficiency. It is to be understood that when HSCU 10 has a given average data rate speed, the burst units DBU-1 through DBU-4 each may have a peak rate greater than the HSCU-10 average data rate.
It is to be understood also that the principles of the present invention can be used in connection with any form of concentrator wherein a high speed unit or line that is overseeing a plurality of burst units and that the buffer arrangement of each of the burst driven ports can be used individually as well as in cooperation with other such units. Further, as the burst units vary in rate, the buffer size and alternation points of the buffers in the various ports may be adapted to the characteristics of the diverse burst units.
While the invention has been particularly shown and described with reference to the preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US2905930 *||May 24, 1954||Sep 22, 1959||Underwood Corp||Data transfer system|
|US3020525 *||Apr 4, 1958||Feb 6, 1962||American Telephone & Telegraph||Record controlled translator|
|US3051929 *||Mar 13, 1959||Aug 28, 1962||Bell Telephone Labor Inc||Digital data converter|
|US3079590 *||Nov 20, 1959||Feb 26, 1963||Gen Dynamics Corp||Data handling system|
|US3209332 *||Sep 7, 1961||Sep 28, 1965||Potter Instrument Co Inc||Reflexing buffer|
|US3293613 *||Apr 5, 1963||Dec 20, 1966||Potter Instrument Co Inc||Information recording system|
|US3344406 *||Nov 9, 1964||Sep 26, 1967||Ibm||Sampled data reduction and storage system|
|US3369223 *||Nov 16, 1965||Feb 13, 1968||Hugh L. Dryden||Incremental tape recorder and data rate converter|
|US3454930 *||Apr 27, 1966||Jul 8, 1969||Potter Instrument Co Inc||Digital magnetic tape recording system|
|US3573745 *||Dec 4, 1968||Apr 6, 1971||Bell Telephone Labor Inc||Group queuing|
|US3588840 *||Feb 5, 1969||Jun 28, 1971||Nomura Takuya||Method of block recording data on a magnetic tape|
|US3631406 *||Nov 19, 1969||Dec 28, 1971||Siemens Ag||Method of continuously exchanging data between a data processing apparatus and external devices|
|US3736568 *||Sep 8, 1971||May 29, 1973||Diginetics Inc||System for producing a magnetically recorded digitally encoded record in response to external signals|
|US3742456 *||Apr 5, 1972||Jun 26, 1973||Pitney Bowes Inc||Apparatus for selectively formatting serial data bits into separate data characters|
|US4003027 *||Oct 4, 1974||Jan 11, 1977||Dynell Electronics Corporation||Arrangement for varying the rate of recording of information|
|US4040027 *||Apr 21, 1976||Aug 2, 1977||U.S. Philips Corporation||Digital data transfer system having delayed information readout from a first memory into a second memory|
|US4087626 *||Aug 4, 1976||May 2, 1978||Rca Corporation||Scrambler and unscrambler for serial data|
|1||*||West et al., "A Digital Computer for Scientific Application", Proceedings of the IRE, Dec. 1948, pp. 1452-1453.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US4344152 *||Nov 29, 1979||Aug 10, 1982||International Business Machines Corp.||Buffer memory control circuit for label scanning system|
|US4433391 *||Aug 17, 1981||Feb 21, 1984||Burroughs Corporation||Buffered handshake bus with transmission and response counters for avoiding receiver overflow|
|US4442485 *||Dec 5, 1980||Apr 10, 1984||Hitachi, Ltd.||Dynamically buffered data transfer system for large capacity data source|
|US4569034 *||Jul 19, 1982||Feb 4, 1986||International Business Machines Corporation||Method and apparatus which allows the working storage to be reconfigured according to demands for processing data input|
|US4733346 *||Aug 8, 1985||Mar 22, 1988||Kabushiki Kaisha Toshiba||Data processor with multiple register blocks|
|US4809161 *||Nov 18, 1985||Feb 28, 1989||Shunichi Torii||Data storage device|
|US4811279 *||Mar 9, 1987||Mar 7, 1989||Digital Equipment Corporation||Secondary storage facility employing serial communications between drive and controller|
|US4945495 *||Oct 20, 1988||Jul 31, 1990||Daikin Industries, Ltd.||Image memory write control apparatus and texture mapping apparatus|
|US5111385 *||Jun 3, 1991||May 5, 1992||Nec Corporation||Parallel-mode data transfer apparatus using sector memories|
|US5179662 *||Aug 31, 1989||Jan 12, 1993||International Business Machines Corporation||Optimized i/o buffers having the ability to increase or decrease in size to meet system requirements|
|US5179665 *||Nov 13, 1989||Jan 12, 1993||Westinghouse Electric Corp.||Microprocessor information exchange with updating of messages by asynchronous processors using assigned and/or available buffers in dual port memory|
|US5230073 *||Jun 6, 1989||Jul 20, 1993||Bell Communications Research, Inc.||System and method for accessing and updating a continuously broadcasted stored database|
|US5295252 *||Jan 13, 1989||Mar 15, 1994||Hitachi, Ltd.||Data storage device|
|US5301274 *||Aug 19, 1991||Apr 5, 1994||Multi-Tech Systems, Inc.||Method and apparatus for automatic balancing of modem resources|
|US5404452 *||Jun 22, 1994||Apr 4, 1995||International Business Machines Corporation||Personal computer bus and video adapter for high performance parallel interface|
|US5404455 *||Dec 31, 1991||Apr 4, 1995||Dictaphone Corporation||Time division multiplexer chip for supporting alternating communication between a pair of RAMs and two different interfaces|
|US5420984 *||Aug 13, 1993||May 30, 1995||Genroco, Inc.||Apparatus and method for rapid switching between control of first and second DMA circuitry to effect rapid switching beween DMA communications|
|US5448701 *||Dec 22, 1992||Sep 5, 1995||International Business Machines Corporation||Flow controller for shared bus used by plural resources|
|US5502822 *||May 14, 1991||Mar 26, 1996||Kabushiki Kaisha Komatsu Seisakusho||Asynchronous data transmission system|
|US5539896 *||Jan 3, 1994||Jul 23, 1996||International Business Machines Corporation||Method and apparatus for dynamically linking code segments in real time in a multiprocessor computing system employing dual buffered shared memory|
|US5644729 *||Jul 28, 1994||Jul 1, 1997||International Business Machines Corporation||Bidirectional data buffer for a bus-to-bus interface unit in a computer system|
|US5655147 *||Apr 19, 1994||Aug 5, 1997||Adaptec, Inc.||SCSI host adapter integrated circuit utilizing a sequencer circuit to control at least one non-data SCSI phase without use of any processor|
|US5671446 *||Mar 16, 1995||Sep 23, 1997||Apple Computer, Inc.||Method and apparatus for atomically accessing a queue in a memory structure where LIFO is converted to FIFO|
|US5696990 *||May 15, 1995||Dec 9, 1997||Nvidia Corporation||Method and apparatus for providing improved flow control for input/output operations in a computer system having a FIFO circuit and an overflow storage area|
|US5717954 *||Oct 13, 1995||Feb 10, 1998||Compaq Computer Corporation||Locked exchange FIFO|
|US5740394 *||Apr 23, 1997||Apr 14, 1998||Mitsubishi Denki Kabushiki Kaisha||Apparatus for transferring data divided into blocks with address boundaries|
|US5745785 *||May 2, 1996||Apr 28, 1998||Sofmap Future Design, Inc.||System for alternatively transferring data from external memory to memory device and from memory device to internal memory depending upon processing unit's operational|
|US5819111 *||Mar 15, 1996||Oct 6, 1998||Adobe Systems, Inc.||System for managing transfer of data by delaying flow controlling of data through the interface controller until the run length encoded data transfer is complete|
|US5904732 *||Apr 30, 1996||May 18, 1999||Sun Microsystems, Inc.||Dynamic priority switching of load and store buffers in superscalar processor|
|US5916309 *||May 12, 1997||Jun 29, 1999||Lexmark International Inc.||System for dynamically determining the size and number of communication buffers based on communication parameters at the beginning of the reception of message|
|US5938748 *||Sep 11, 1997||Aug 17, 1999||Apple Computer, Inc.||Data transfer mechanism for transfer of real-time data including real-time data computed on-the-fly|
|US6031624 *||Sep 8, 1997||Feb 29, 2000||Lexmark International, Inc.||Method and apparatus for adaptive data buffering in a parallelized printing system|
|US6046817 *||Jan 22, 1998||Apr 4, 2000||Lexmark International, Inc.||Method and apparatus for dynamic buffering of input/output ports used for receiving and transmitting print data at a printer|
|US6067408 *||Feb 22, 1996||May 23, 2000||Advanced Micro Devices, Inc.||Full duplex buffer management and apparatus|
|US6088716 *||Apr 28, 1997||Jul 11, 2000||Ab Initio Software Corporation||Method for preventing buffer deadlock in dataflow computations|
|US6170077 *||Aug 18, 1998||Jan 2, 2001||Electronics And Telecommunications Research Institute||Method for encoding a digital communication channel|
|US6246752||Jun 8, 1999||Jun 12, 2001||Valerie Bscheider||System and method for data recording|
|US6249570||Jun 8, 1999||Jun 19, 2001||David A. Glowny||System and method for recording and storing telephone call information|
|US6252946||Jun 8, 1999||Jun 26, 2001||David A. Glowny||System and method for integrating call record information|
|US6252947||Jun 8, 1999||Jun 26, 2001||David A. Diamond||System and method for data recording and playback|
|US6311265||Mar 25, 1996||Oct 30, 2001||Torrent Systems, Inc.||Apparatuses and methods for programming parallel computers|
|US6330008||Feb 24, 1997||Dec 11, 2001||Torrent Systems, Inc.||Apparatuses and methods for monitoring performance of parallel computing|
|US6445700 *||Mar 31, 1998||Sep 3, 2002||Oki Electric Industry Co., Ltd.||Serial communication circuit|
|US6697904 *||Mar 28, 2000||Feb 24, 2004||Intel Corporation||Preventing starvation of agents on a bus bridge|
|US6728345 *||Jun 8, 2001||Apr 27, 2004||Dictaphone Corporation||System and method for recording and storing telephone call information|
|US6754740||Dec 11, 2000||Jun 22, 2004||Nokia Corporation||Interface apparatus for connecting devices operating at different clock rates, and a method of operating the interface|
|US6775372||Jun 2, 1999||Aug 10, 2004||Dictaphone Corporation||System and method for multi-stage data logging|
|US6785369 *||Jun 8, 2001||Aug 31, 2004||Dictaphone Corporation||System and method for data recording and playback|
|US6874043||Apr 22, 2002||Mar 29, 2005||Bridgeworks Ltd.||Data buffer|
|US6937706 *||Jun 8, 2001||Aug 30, 2005||Dictaphone Corporation||System and method for data recording|
|US7000139 *||Aug 28, 2002||Feb 14, 2006||Oki Electric Industry Co., Ltd.||Interface circuit for selectively latching between different sets of address and data registers based on the transitions of a frequency-divided clock|
|US7065596||Sep 19, 2002||Jun 20, 2006||Intel Corporation||Method and apparatus to resolve instruction starvation|
|US7805551 *||Sep 27, 2007||Sep 28, 2010||Integrated Device Technology, Inc.||Multi-function queue to support data offload, protocol translation and pass-through FIFO|
|US7818470||Sep 27, 2007||Oct 19, 2010||Integrated Device Technology, Inc.||Adaptive interrupt on serial rapid input/output (SRIO) endpoint|
|US7945716||Sep 27, 2007||May 17, 2011||Integrated Device Technology, Inc.||Serial buffer supporting virtual queue to physical memory mapping|
|US8359117 *||Dec 28, 2009||Jan 22, 2013||Hitachi Kokusai Electric Inc.||Substrate processing system|
|US8489783 *||Jan 3, 2007||Jul 16, 2013||Apple Inc.||Multi buffer asynchronous scheme for processing incoming information|
|US20010040942 *||Jun 8, 2001||Nov 15, 2001||Dictaphone Corporation||System and method for recording and storing telephone call information|
|US20010043685 *||Jun 8, 2001||Nov 22, 2001||Dictaphone Corporation||System and method for data recording|
|US20010055372 *||Jun 8, 2001||Dec 27, 2001||Dictaphone Corporation||System and method for integrating call record information|
|US20020080181 *||Nov 20, 2001||Jun 27, 2002||Razdow Allen M.||Apparatuses and methods for monitoring performance of parallel computing|
|US20040068597 *||Sep 19, 2002||Apr 8, 2004||Kulick S. Steven||Method and apparatus to resolve instruction starvation|
|US20040106017 *||Oct 22, 2001||Jun 3, 2004||Harry Buhay||Method of making coated articles and coated articles made thereby|
|US20100199001 *||Aug 5, 2010||Hitachi-Kokusai Electric, Inc.||Substrate processing system|
|EP0099462A2 *||Jun 8, 1983||Feb 1, 1984||International Business Machines Corporation||Apparatus and method for buffering data in a data processing system|
|EP0262468A2 *||Sep 7, 1987||Apr 6, 1988||Advanced Micro Devices, Inc.||Reconfigurable fifo memory device|
|EP0415862A2 *||Jun 19, 1990||Mar 6, 1991||International Business Machines Corporation||Optimized I/O buffers|
|WO1998049628A2 *||Apr 28, 1998||Nov 5, 1998||Initio Software Corp Ab||Method for preventing buffer deadlock in dataflow computations|
|WO1999066392A1 *||Jun 17, 1998||Dec 23, 1999||Happonen Aki||An interface apparatus for connecting devices operating at different clock rates, and a method of operating the interface|
|WO2008085787A2 *||Dec 28, 2007||Jul 17, 2008||Apple Inc||Multi buffer asynchronous scheme for processing incoming information|
|WO2013101581A1 *||Dec 19, 2012||Jul 4, 2013||Schlumberger Canada Limited||Inter-tool communication flow control in toolbus system of cable telemetry|
|International Classification||G06F13/38, G06F13/00, G06F5/16, G06F5/06, G06F3/06|